US 6141749

I'll search for authoritative information on this specific patent number, including any litigation.

Let me probe specifically for litigation/docket activity and reassignment records.

US Patent 6,141,749 — Summary

Caveat up front: I retrieved this information from Google Patents (which mirrors USPTO bibliographic data), FreePatentsOnline, and uspto.report. I did not obtain direct USPTO PatentCenter or PACER/CAFC docket access in this session. See "Litigation/docket status" below for what I could and could not confirm.

Bibliographic data

Abstract (as issued)

The abstract states that the invention provides improved computer network firewalls with features for increased processing efficiency: support for multiple security policies and/or multiple users by applying any one of several distinct sets of access rules; "stateful" packet filtering involving caching rule‑processing results and using the cached results to bypass rule processing for subsequent similar packets; a dependency mask set from session data items (source host address, destination host address, type of service) usable to query a cache of active sessions so a rule can be selected based on the number of matching sessions; dynamic rules used in addition to pre‑loaded access rules; and the ability to redirect a network session to a separate server for processing to unburden the firewall of application proxies.

Plain-language overview of the independent claims

All four independent claims are directed to the stateful packet filtering aspect (the second aspect in the specification). Each independent claim has the same three substantive elements, and the four are essentially parallel method / apparatus / means‑plus‑function / programmed‑system forms.

• Claim 1 — Method for packet validation in a computer network firewall. Three steps:

  1. Store in a cache (i) a result of applying at least a portion of a rule set to a given packet of a network session, and (ii) at least one hardware address associated with that given packet.
  2. Use the stored result to process at least one subsequent packet that is similar to the given packet, where a session key associated with the subsequent packet is used to retrieve the stored result from the cache.
  3. Determine the destination of the subsequent packet — depending on the result of the processing — using the stored hardware address.

  The distinguishing feature versus a generic "cache the firewall decision" concept is element (3): the destination determination for the follow‑on packet is made using the hardware address cached alongside the rule result.

• Claim 5 — Apparatus for validating a packet in a firewall: a memory storing a cache containing the same two items as claim 1 (rule‑processing result + hardware address), and a processor coupled to the memory operative to perform the same utilizing and destination‑determining steps.

• Claim 9 — Computer system expressed in means‑plus‑function format: means for storing the cache contents, means for utilizing the stored result with the session key, and means for determining the destination using the stored hardware address.

• Claim 13 — Computer system comprising a processor instructed for the same storing, utilizing, and destination‑determining operations.

Dependent claims (all four families mirror each other):

• Claims 2, 6, 10, 14: the utilizing step/means is further operative to bypass rule processing for the subsequent packet.
• Claims 3, 7, 11, 15: the "similar characteristic" is that the packets are from the same network session.
• Claims 4, 8, 12, 16: the stored hardware address is a local area network (LAN) address.

Scope note / potential §112 issue: Claim 9 is drafted in means‑plus‑function language, and claim 13 recites a processor "instructed for" steps rather than a programmed processor with stored instructions — the kind of phrasing that historically invited indefiniteness challenges (cf. the related Sound View portfolio patents where the PTAB declined institution on the ground that asserted means‑plus‑function claims lacked sufficient structure). I have not seen a PTAB or court ruling specifically construing claim 9 or 13 of the '749 patent.

Notable cited prior art (from the face of the patent)

US 5,349,642 (Kingdon); US 5,699,513 (Feigen et al.); US 5,706,427 (Tabuki); US 5,826,014 (Coley et al.); US 5,828,833 (Belville et al.); US 5,842,040 (Hughes et al., Storage Technology Corp., "Policy caching method…"); US 5,848,233 (Radia et al.); EP 0 743 777 A2; EP 0 856 974 A2 (AT&T, "Session cache and rule caching method for a dynamic filter"); WO 97/00471; WO 97/02734; WO 97/49038. The most directly material references appear to be the Storage Technology policy‑caching patent and the AT&T session/rule‑caching application.

Litigation / docket status

• I found no evidence of any CAFC 2026 docket involving 6,141,749, and no district court or PTAB proceeding asserting this specific patent. Multiple searches for the number plus "CAFC 2026," "Federal Circuit appeal," and "IPR petition" returned nothing tied to 6,141,749.
• Assignee litigation context: Sound View Innovations, LLC asserted a large Lucent‑derived portfolio (e.g., 5,806,062; 5,847,972; 5,918,009; 6,125,371; 6,240,391; 6,408,296; 6,502,133; 6,725,456; 7,412,486; 7,894,859; 8,095,593; 8,135,860) against Facebook, LinkedIn, Hulu, Walmart, Delta, Cigna, and DISH/Sling. Those suits and IPRs (e.g., IPR2017‑00985 through ‑01006; IPR2018‑00096; IPR2020‑01041) target other patents in the portfolio, not 6,141,749.
• Practical implication: because the '749 patent's 20‑year term ran from September 12, 1997 and it is listed as expired as of 2017-09-12, a new 2026 appeal on this patent would be unusual, though an old appeal could theoretically still be resolved.
• Uncertainty: I cannot affirmatively certify that there is zero docket activity. Without PACER/CAFC CM‑ECF access I cannot exclude a sealed, unindexed, or very recent filing. Treat the "no 2026 CAFC activity" finding as based on indexed web sources only.

One data discrepancy to flag

The Unified Patents portal lists the '749 priority date as 1997-09-11, whereas Google Patents, uspto.report, FreePatentsOnline, and the EP family member (EP 98306984.0, priority "12.09.1997 US 928794") all say 1997-09-12. I am treating September 12, 1997 as correct per the authoritative patent text provided, and noting the 9/11 entry as a likely data error in that third‑party record rather than auto‑correcting it silently.

I'll research litigation involving US patent 6141749. Let me search multiple sources.

Let me search more specifically for litigation records.

Let me search for specific litigation and PTAB proceedings.

Summary of Findings

I could not identify any litigation involving U.S. Patent No. 6,141,749 from the searches performed. Based on the sources searched, there does not appear to be any publicly recorded litigation (district court or Federal Circuit) specifically asserting this patent.

Searches performed

I searched the web (including Unified Patents' patent portal results, Justia, FreePatentsOnline, USPTO.report, and PTAB/docket-associated documents) for the exact number 6141749 / 6,141,749 combined with litigation, infringement, IPR, and NPE terms. The searches returned:

• The patent's own documents (Google Patents, Justia, FreePatentsOnline) and citation/citing-patent lists — no litigation records.
• Numerous Sound View Innovations, LLC cases in the District of Delaware and Central District of California — but the asserted patents in those matters were different numbers (e.g., 5,806,062; 6,125,371; 6,708,213; 6,757,796; 9,462,074; 7,426,715; 6,502,133; 6,725,456). None of those records referenced 6141749.

No PACER docket, Federal Circuit appeal, or PTAB (IPR/PGR) record tied to 6141749 surfaced.

Important caveat

This is a negative finding based on limited search coverage, not a certification that no litigation exists. Web search is not exhaustive for district court dockets, and PACER/Unified Patents litigation databases were not directly queried in full. To be certain, you should verify directly via:

• Unified Patents litigation portal – https://portal.unifiedpatents.com/litigation/caselist (filter by patent number)
• PACER – search "6,141,749" across district courts
• USPTO PTAB – search patent number on the PTAB Endpoint/Patent Trial and Appeal Board docket
• CourtListener / RECAP – https://www.courtlistener.com/?q=%226%2C141%2C749%22

Relevant context on the patent itself (grounded in the record I retrieved)

• Patent: U.S. 6,141,749 — "Methods and apparatus for a computer network firewall with stateful packet filtering"
• Inventors: Michael John Coss, David L. Majette, Ronald L. Sharp
• Original assignee: Lucent Technologies Inc.
• Ownership chain: Lucent Technologies → Alcatel-Lucent USA Inc. → Sound View Innovations, LLC (assignment recorded 2014-07-25) → and later reassignments back toward Nokia of America / Alcatel Lucent.
• Legal status: Expired – Lifetime; anticipated expiration 2017-09-12.
• Family: EP0909072A3 (withdrawn) and JP3464610B2 (expired).

Sound View Innovations is a well-known non-practicing entity that filed many patent lawsuits (e.g., against Hulu, Netflix-adjacent streaming/retail defendants, Walmart/Vudu, Delta, Cigna, HSN, QVC, AMC, CBS, Showtime), but on the record retrieved, 6,141,749 was not among the patents those cases asserted.

Bottom line

• Known litigation involving U.S. 6,141,749: none found.
• No plaintiff, defendant, jurisdiction, case number, filing date, or outcome can be reported, because no case was located.
• Given the patent is expired (since 2017), the likelihood of new assertions is low, though historical litigation records should be confirmed through the docket-specific sources listed above before relying on this negative result.

If you can point me to a specific court or party you suspect, I can run a narrower search.

I'll start by checking the canonical structured data against live sources, then search for any proceedings the ODP may not have indexed.

Let me drill into whether the '749 patent itself was ever challenged or asserted.

Let me check the defensive-aggregator and PTAB-specialist databases directly.

PTAB Proceedings Report — U.S. Patent No. 6,141,749

Patent: US 6,141,749 A ("Methods and apparatus for a computer network firewall with stateful packet filtering")
Application: US 08/928,794 | Filed: 1997-09-12 | Granted: 2000-10-31
Inventors: Michael John Coss, David L. Majette, Ronald L. Sharp
Original assignee: Lucent Technologies Inc. | Assignment chain (per patent face): Lucent → Alcatel-Lucent USA → Sound View Innovations, LLC (assigned 2014-07-25) → with later 2019 nunc pro tunc assignments recorded to Nokia of America Corp./Alcatel Lucent
Statutory term: Expired 2017-09-12 (20-year term; Google Patents status "Expired – Lifetime")

---

Proceedings overview

Total AIA trial proceedings on file for US 6,141,749: zero. The USPTO Open Data Portal (ODP) structured "PTAB proceedings on file" block returns no AIA trial proceedings, and independent web searches of PTAB-related sources for this patent number returned no IPR, PGR, or CBM — so the breakdown is 0 active / 0 claims-invalidated / 0 claims-sustained / 0 settled / 0 institution-denied, and the defensive posture this gives a defendant is the strongest kind: no PTAB petitioner has ever attacked this patent, no claim has ever been canceled, and the patent's statutory term expired on 2017-09-12 — meaning there is no forward-looking assertion to defend against and no six-year § 286 damages window still reachable for a complaint filed today (2026-09-30).

Methodological note and caveat. The ODP block is the canonical list and it is empty. My web searches were run to surface anything ODP had not indexed. They surfaced extensive Sound View Innovations PTAB activity on sibling Lucent/Bell Labs patents — US 6,125,371 ('371), US 6,725,456 ('456), US 6,732,181 ('181), US 7,366,786 ('786), US 5,991,845 ('845), and others — but nothing on the '749 itself. I cannot rule out a non-indexed or very recent filing, but the absence is consistent across the ODP data and every secondary source I checked.

No proceedings to report

There are no PTAB proceeding numbers to list, and per the operating constraints I will not invent any. To avoid the common research error of confusing this patent's family with its siblings, the table below lists the adjacent Sound View PTAB matters I actually encountered — none of these is an IPR/PGR/CBM of the '749, and none of them invalidates, narrows, or estops anything with respect to the '749.

Defensive value from the '749's own record: because no petitioner ever filed, there is no § 315(e)(2) estoppel, no adverse FWD, no claim construction from the Board, and no canceled claim. A demand letter citing any of claims 1–16 has to be evaluated purely on the patent's own footing — which, as set out below, is a footing the patent largely lost to the calendar.

---

Strategic summary

Which claims are canceled vs. sustained vs. untested. All 16 claims of the '749 (independent claims 1, 5, 9, and 13, plus dependent claims 2–4, 6–8, 10–12, and 14–16) are UNTESTED at the PTAB. None was canceled, none was confirmed, none was construed by the Board. That is a genuine fact but not a favorable one for the patent owner: the claims are untested and the patent is expired, so there is no live controversy in which they could be tested. The claim set breaks down as independent claim 1 (method for packet validation with cached rule-set result + hardware address), independent claim 5 (apparatus with memory + processor), independent claim 9 (computer system in means-plus-function form for storing/utilizing/determining), and independent claim 13 (computer system with a processor "instructed for" the same steps). Interesting defensive point: claims 9 and 13's near-verbatim overlap with claims 1 and 5 means a § 112 indefiniteness or written-description attack on the means-plus-function claim 9 — the tactic Facebook used successfully against Sound View's '181 and '786 patents, where the Board declined institution only because the asserted means-plus-function claims lacked sufficient structure — remains available if anyone ever revives an assertion. See the Facebook stay briefing record, D. Del. C.A. No. 16-116, D.I. 201.

Estoppel landscape. There is no estoppel on this patent. § 315(e)(2) estops only a petitioner that obtained an FWD; no petitioner exists here, so no prior-art ground is barred against anyone. Any defendant is free to run any § 102/§ 103/§ 112 ground it can develop, without worrying that a prior challenger's IPR already locked in a subset. Conversely, the patent owner has never had to defend these claims and has no Board or Federal Circuit rulings to rely on for a presumption of validity boost. Practically, though, the more decisive point for a defendant is not estoppel but expiration: with the term ending 2017-09-12, a complaint served on 2026-09-30 can only reach back six years under 35 U.S.C. § 286 — to 2020-09-30 — which is entirely after the patent expired. Absent an earlier complaint or a tolling agreement running from a pre-2017 filing, there is no recoverable infringement period, which is a complete § 286/standing-type defense that obviates the need for an IPR. (Caveat: I could not rule out a pre-2017 assertion campaign on the '749 from the sources I reviewed; that is a docket-level diligence item, see next steps.)

Pattern signals. No petitioner has filed on the '749, let alone multiple petitions — so none of the "serial IPR" patterns that mark a heavily asserted patent. The patent's ownership history is the notable signal: it sits in the Sound View Innovations portfolio — the NPE that acquired 1,000+ Lucent/Alcatel-Lucent patents in December 2013 and, per contemporaneous reporting, filed 13 infringement suits from 2016–2019 (Facebook, Hulu, Twitter, Fidelity, Walmart, Delta, AMC, CBS, The CW, DISH). That campaign drew sustained PTAB fire from sophisticated defendants and from the defensive aggregator Unified Patents, LLC (which both filed IPRs and requested the ex parte reexam on the '456). The signal to draw from the '749's complete absence from that campaign is that the patent owner itself does not appear to have treated the '749 as a monetizable asset. The 2019 nunc pro tunc assignment records moving the patent from Nokia of America Corp. to Alcatel Lucent are also worth noting — on their face they suggest a reversion or correction that may have taken the '749 out of Sound View's enforcement pool; I could not confirm the current owner with high confidence from the pages I retrieved.

---

Recommended next steps

1. Do not file an IPR — the patent is expired and there is nothing left to invalidate for practical purposes. A petition on an expired patent is technically permissible (the Board has instituted on expired patents, and the owner can no longer amend), but it would cost a full IPR budget to defeat a claim that cannot support forward relief and, for a 2026 complaint, appears to support no § 286 damages window at all. Confirm the term calculation first: application US 08/928,794 was filed 1997-09-12, so the 20-year term under 35 U.S.C. § 154 ran to 2017-09-12, consistent with the Google Patents docket entry "2017-09-12 Anticipated expiration" and status "Expired – Lifetime." Check maintenance-fee history too — an earlier lapse (with no § 41(c) petition to revive) would push the damages cutoff even further into the past.

2. Run the § 286 / assertion-history diligence. Pull the district court dockets and confirm whether the '749 was ever asserted before expiration and whether any earlier complaint or tolling agreement could pull recoverable acts into the pre-2017 window. The Sound View campaign complaints I reviewed asserted '062, '972, '009, '391, '296, '859, '133, '845, '371, '181, '786, '456, '213, '074, and '715 — but not the '749. If no prior assertion exists, the expiration defense is likely dispositive and no invalidity workup is needed.

3. Confirm current ownership before responding to any demand. The Google Patents assignment record shows Sound View Innovations, LLC taking assignment on 2014-07-25, followed by 2019-09-24 and 2019-10-08 nunc pro tunc assignments involving Nokia of America Corporation and Alcatel Lucent. A demand letter from an entity that does not hold title is a standing problem, and a reverted patent may mean the asserting party lacks rights. Verify on the USPTO Assignment Search and Patent Center, and confirm the patent is not the subject of a post-expiration terminal disclaimer or abandoned litigation.

4. If the patent owner nonetheless asserts, lead with the calendar, not the prior art. The strongest available responses, in order, are (a) expiration on 2017-09-12 with no reachable § 286 damages period for a 2026 filing; (b) no forward-looking relief available for an expired patent; (c) § 287 marking/failure-to-mark and laches-type limitations on any older theory; and (d) — only if a pre-2017 damages window somehow survives — § 112 indefiniteness on the means-plus-function claim 9 and the parallel claims 9/13, the exact theory that kept Sound View's '181 and '786 claims out of the PTAB.

5. If you want the absence of IPRs on the record, cite the primary sources directly.

• USPTO ODP / PTAB structured data: https://developer.uspto.gov/ptab-api
• PTAB E2E / PTAB Center (proceeding search by patent number): https://ptacts.uspto.gov/
• Patent face and legal-status docket: https://patents.google.com/patent/US6141749/en
• Federal Circuit opinions (for the sibling-patent appeals, if useful context): https://www.courtlistener.com — e.g., DISH Network L.L.C. v. Sound View Innovations, LLC, No. 22-1627 (Fed. Cir. Oct. 5, 2023) (affirming IPR2020-01041 FWD), https://www.courtlistener.com/opinion/9430285/dish-network-llc-v-sound-view-innovations-llc/

---

Bottom line for a defendant

There is no PTAB activity on file for US 6,141,749 — zero IPRs, zero PGRs (unavailable anyway for a 1997 pre-AIA filing), zero CBMs. Nothing has been canceled, so a demand citing claims 1–16 is not sanction-bait; but the patent expired on 2017-09-12 and has never been used in the Sound View assertion campaign, which means the practical answer to a demand letter is a term/§ 286 response rather than an invalidity budget. Verify title and the maintenance-fee and assertion history first; if those check out, the absence of PTAB proceedings is not a gap in the record — it is a signal that nobody has ever thought these claims worth challenging.

Flagged uncertainty: (i) I relied on the ODP structured block for the canonical "no proceedings" finding and corroborated it by search; I could not access an authoritative live PTAB API response in this session; (ii) the current owner of record is ambiguous from the assignment chain on the patent face; (iii) I did not exhaustively enumerate every district court docket in which the '749 may have been mentioned.

I'll research the assignment record for US 6,141,749, focusing on reel/frame data, correspondents of record, and the Sound View Innovations chain.

Let me dig into the specific reel/frame records and correspondents.

I hit the tool-call limit before I could pull the Assignment Center reel/frame detail directly, so I'll write up what the records I did retrieve support and flag the gaps explicitly rather than fill them by inference.

Data-source caveat (read first): I could not reach assignmentcenter.uspto.gov / assignment.uspto.gov directly in this session, and my searches for the specific reel/frame strings returned nothing. The assignment chain below is reconstructed from the Google Patents legal-events table in the authoritative patent text provided, which mirrors USPTO assignment records but suppresses the reel/frame, execution date, and correspondent-of-record fields. Every reel/frame and correspondent value that I could not verify is marked [not retrieved — not fabricated]. Related-chain correspondents I did surface (from a different Alcatel-Lucent application) are labeled as such and are not asserted to be correspondents on '749.

---

Inventors

• All three are named on the inventor→Lucent ASSIGNMENT OF ASSIGNORS INTEREST recorded 1998-03-11, i.e. a standard employee invention assignment to the corporate assignee. There is no inventor-held assignment, no subsequent inventor reassignment, and no evidence any inventor was an assignee of record after 1998.
• Pattern note: the classic "inventors peel off and the portfolio gets sold" tell is not present at the inventor level. The monetization event that matters for this patent happened 17 years later through the corporate chain (Lucent → Alcatel-Lucent USA → Sound View), not through inventor activity. I could not confirm any inventor's post-Lucent employment from indexed sources.
• Residences are all within a ~30-mile radius of Lucent's Murray Hill / NJ Bell Labs campus, consistent with a single Bell Labs team.

---

Original assignee

Lucent Technologies Inc. (P.O. Box 636, 600 Mountain Avenue, Murray Hill, NJ 07974) — named on the issued patent.

• Line of business: telecommunications and networking equipment/systems vendor, spun out of AT&T in 1996, carrying AT&T Bell Laboratories R&D.
• Did they ship a product embodying the claims? [unclear / not confirmed.] The specification states prototypes were implemented in C on general-purpose PC hardware, which indicates active internal development, but I found no indexed evidence of a specific Lucent commercial firewall product shipping these claims. I will not assert either way.
• Current status: operating entity absorbed via merger. Lucent Technologies Inc. → Alcatel-Lucent USA Inc. (MERGER recorded 2014-05-29) → renamed Nokia of America Corporation (CHANGE OF NAME recorded 2019-09-24) after Nokia's 2016 acquisition of Alcatel-Lucent. So the corporate lineage of the original assignee now sits inside Nokia. The patent itself left that lineage in 2014 (see below).

---

Assignment timeline

All dates below are recordation dates from the Google Patents legal-events table. Execution dates are not exposed by that source and I did not retrieve them.

• recorded 1998-03-11 — Reel [not retrieved — not fabricated]

  • Conveyance: Assignment of assignors' interest
  • Assignor: Coss, Michael John; Majette, David L.; Sharp, Ronald L.
  • Assignee: Lucent Technologies Inc.
  • Correspondent: [not retrieved]
  • Context: standard employee→employer invention assignment at filing.

• recorded 2001-04-05 — Reel [not retrieved — not fabricated]

  • Conveyance: Conditional assignment and security interest in patent rights
  • Assignor: Lucent Technologies Inc. (DE corporation)
  • Assignee: The Chase Manhattan Bank, as collateral agent
  • Correspondent: [not retrieved]
  • Context: securitization/collateral — Lucent pledged patent rights to secure debt; not a change of beneficial ownership.

• recorded 2006-12-06 — Reel [not retrieved — not fabricated]

  • Conveyance: Termination and release of security interest in patent rights
  • Assignor: JPMorgan Chase Bank, N.A. (f/k/a The Chase Manhattan Bank), as administrative agent
  • Assignee: Lucent Technologies Inc.
  • Correspondent: [not retrieved]
  • Context: release of the 2001 collateral lien — internal/securitization cleanup, no ownership change.

• recorded 2013-03-07 — Reel [not retrieved — not fabricated]

  • Conveyance: Security interest
  • Assignor: Alcatel-Lucent USA Inc.
  • Assignee: Credit Suisse AG
  • Correspondent: [not retrieved] — related-chain note: in a parallel Alcatel-Lucent application (14264538), the Credit Suisse security-interest/release recordings were handled by Kirkland & Ellis LLP (Edward Sadler / Hayley Smith Sr.), New York. Not confirmed for '749.
  • Context: securitization/collateral — new lender lien on the Alcatel-Lucent patent estate (the estate that was about to be monetized). Not a change of beneficial ownership.

• recorded 2014-05-29 — Reel [not retrieved — not fabricated]

  • Conveyance: Merger
  • Assignor: Lucent Technologies Inc.
  • Assignee: Alcatel-Lucent USA Inc.
  • Correspondent: [not retrieved]
  • Context: internal corporate reorganization — Lucent folded into Alcatel-Lucent USA; title housekeeping, no third-party transfer.

• recorded 2014-07-25 — Reel [not retrieved — not fabricated]

  • Conveyance: Assignment of assignors' interest
  • Assignor: Alcatel Lucent
  • Assignee: Sound View Innovations, LLC
  • Correspondent: [not retrieved] — related-chain note: a same-era Alcatel-Lucent→WSOU Investments, LLC recording (application 14264538, reel 033655/0304) lists correspondent Sean D. Burdick, 2537 W. State Street, Suite 220, Boise, ID 83702, with the assignee at 11150 Santa Monica Blvd., Suite 1400, Los Angeles, CA. WSOU and Sound View are sibling Alcatel-Lucent-spinout NPEs; I could not confirm Burdick (or any correspondent) on the '749 recording.
  • Context: transfer-to-asserter — the monetization event. This is the load-bearing assignment for NPE status.

• recorded 2014-10-09 — Reel [not retrieved — not fabricated]

  • Conveyance: Release by secured party
  • Assignor: Credit Suisse AG
  • Assignee: Alcatel-Lucent USA Inc.
  • Correspondent: [not retrieved] — related-chain note: Kirkland & Ellis appears on the corresponding Credit Suisse release in the parallel application.
  • Context: securitization cleanup — lien released after the estate was sold off.

• recorded 2019-09-24 — Reel [not retrieved — not fabricated]

  • Conveyance: Change of name
  • Assignor: Alcatel-Lucent USA Inc.
  • Assignee: Nokia of America Corporation
  • Correspondent: [not retrieved]
  • Context: change of name only — Alcatel-Lucent USA renamed after Nokia's acquisition.

• recorded 2019-10-08 — Reel [not retrieved — not fabricated]

  • Conveyance: Nunc pro tunc assignment
  • Assignor: Nokia of America Corporation
  • Assignee: Alcatel Lucent (Paris)
  • Correspondent: [not retrieved]
  • Context: internal reorg/correction — retroactive title housekeeping on the Alcatel-Lucent side.

⚠️ Contradiction to flag (do not smooth over)

The Google Patents "Current Assignee" field lists both Alcatel Lucent SAS and Sound View Innovations, LLC, and the chain contains two mutually inconsistent ownership threads:

• NPE thread: Alcatel Lucent → Sound View Innovations, LLC (recorded 2014-07-25), consistent with the independently documented Sound View acquisition of ~250 assets (Dec 2013) and ~300 assets (mid-2014).
• Alcatel/Nokia thread: Credit Suisse release (2014-10-09), Change of Name to Nokia of America (2019-09-24), and a nunc pro tunc assignment to Alcatel Lucent (2019-10-08).

The 2019 Alcatel/Nokia entries look like administrative title cleaning on the vendor side, not a reversion of the Sound View grant — but "nunc pro tunc" language in particular is a correction device, and I cannot resolve from the legal-events table alone whether these 2019 records were scoped to patents not sold to Sound View. Resolving it requires reading the actual reel/frame instruments, which I did not retrieve. This is the single most important thing to verify against Assignment Center.

---

Timeline diagram

timeline
    title Ownership of US 6141749
    1997 : Filed by Lucent Technologies
    1998 : Inventors assign to Lucent
    2000 : Patent issued
    2001 : Security interest to Chase Manhattan
    2006 : Security interest terminated and released
    2013 : Credit Suisse security interest recorded
         : Sound View Innovations formed in Delaware
    2014 : Lucent merges into Alcatel-Lucent USA
         : Alcatel Lucent assigns to Sound View Innovations
         : Credit Suisse releases security interest
    2016 : Sound View files first infringement suits
    2017 : Patent term expires
    2019 : Alcatel-Lucent USA renamed Nokia of America
         : Nunc pro tunc assignment to Alcatel Lucent

---

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT.
Patent moved from an operating telecom vendor (Alcatel Lucent) to Sound View Innovations, LLC by ASSIGNMENT OF ASSIGNORS' INTEREST recorded 2014-07-25. Corroborated (independent of the '749 record) by RPX: Sound View was "formed in Delaware in November 2013" and acquired over 250 assets within a month, then nearly 300 more in June 2014. Its recorded address is a waterfront office-suite commercial address, not a corporate HQ — 2001 Route 46, Waterview Plaza, Suite 310, Parsippany, NJ 07054 (per INPI's EP1719282 record naming Sound View as titulaire). Name suffix "Innovations, LLC," no products in commerce, licensing-only. Reel/frame [not retrieved] — verify at Assignment Center.

2. Known asserter in the chain — PRESENT.
Sound View Innovations, LLC is named as an NPE/PAE by Unified Patents ("a well known NPE," in its IPR2018-00096 commentary) and classified "NPE (Patent Assertion Entity)" in Unified's litigation portal (e.g., 1:19-cv-03707/03709 against DISH/Sling, 1:19-cv-00660 against Walmart/Vudu, 1:19-cv-00659 against Delta, 1:19-cv-00964 against Cigna, 1:16-cv-00116 against Facebook). RPX documents the same first-suit wave. Important scoping caveat: Sound View asserted other patents from the same Alcatel-Lucent portfolio (5,991,845; 6,125,371; 6,732,181; 7,366,786; 7,412,486; 8,095,593; 8,135,860, etc.). I found no assertion of US 6,141,749 itself. So '749 is a portfolio asset of a confirmed NPE, not a litigated patent.

3. Repeat correspondent across the chain — UNCLEAR (not established for this patent).
The '749 legal-events record exposes no correspondent field, so I cannot demonstrate recurrence on this chain. For transparency, the correspondents that recur on the related Alcatel-Lucent securitization/monetization paperwork are Kirkland & Ellis LLP (Credit Suisse security-interest and release recordings) and — on a sibling Alcatel-Lucent→WSOU Investments recording — Sean D. Burdick (Boise, ID). These are exactly the type of repeat-player agent the signal is meant to catch, but applying them to '749 would be inference, not evidence. Pull the reel/frame correspondent fields before treating this as a finding.

4. Cascading transfers — PRESENT (with a caveat).
Four recorded events fall inside a ~19-month window: 2013-03-07 (Credit Suisse security interest) → 2014-05-29 (Lucent/Alcatel-Lucent merger) → 2014-07-25 (→ Sound View) → 2014-10-09 (Credit Suisse release). That is <24 months of consecutive changes of record. Caveat: three of the four are securitization/merger mechanics, not independent third-party sales — only the 2014-07-25 event is a genuine transfer-to-asserter. I score this present because the cluster nonetheless shows the portfolio being carved up and liened immediately before monetization, but the evidentiary weight rests on the single Sound View entry.

5. Pre-litigation transfer — NOT PRESENT (for this patent).
Sound View's first asserted suit was filed 2016-03 (per RPX); the '749 transfer was 2014-07-25 — roughly 19 months earlier, outside the 6-month window, and '749 was not among the asserted patents. No evidence the '749 chain was arranged to enable a specific assertion.

6. Bankruptcy fire-sale — NOT PRESENT.
No Chapter 7/11 for Lucent Technologies, Alcatel-Lucent, or Alcatel-Lucent USA in this record. Alcatel-Lucent monetized the estate and was later acquired by Nokia (a strategic acquisition, not a liquidation). The '749 chain is a portfolio divestiture, not a bankruptcy sale.

7. Privateering — PRESENT.
IAM Media covered the Sound View deal under the headline "A hook-up with team behind Micron NPE Round Rock is Alcatel-Lucent's latest foray into IP sales" (Feb 14, 2014), and Sound View's own release states it "acquired approximately 500 Alcatel-Lucent patents in force worldwide." That is the textbook privateering shape: an operating vendor (Alcatel-Lucent) seeds a licensing vehicle with ~500 patents, which then asserts them against operating companies — Sound View sued Facebook, Hulu, Fidelity, Walmart/Vudu, Delta, Cigna, DISH/Sling, HSN, and QVC. Whether Alcatel-Lucent retained a revenue share is not established by the records I retrieved.

8. Defensive aggregator — NOT PRESENT.
The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. It terminates at Sound View Innovations, LLC — the asserter itself.

---

Verdict

NPE — high confidence.

Three strong signals converge: (a) shell-entity transfer — Alcatel Lucent → Sound View Innovations, LLC, a Delaware LLC formed November 2013, ASSIGNMENT OF ASSIGNORS' INTEREST recorded 2014-07-25, at a commercial office-suite address with no products; (b) known asserter — Sound View is designated an NPE/PAE by Unified Patents and its litigation portal and is documented by RPX as acquiring ~250 + ~300 Alcatel-Lucent assets; and (c) privateering — IAM Media explicitly framed the transfer as Alcatel-Lucent's patent-monetization play tied to the team behind the Round Rock NPE. The patent is a portfolio asset of a confirmed NPE rather than a litigated assertion vehicle: I found no suit or PTAB proceeding naming US 6,141,749, and it expired 2017-09-12, so its assertion value dissipated before Sound View's litigation campaign peaked.

Two open items that could change the picture, both requiring the actual reel/frame instruments:

1. The 2019 Alcatel/Nokia records (Change of Name 2019-09-24; nunc pro tunc assignment to Alcatel Lucent 2019-10-08) coexist in the chain with the 2014 Sound View grant — verify their scope.
2. Reel/frame numbers and correspondents were not retrievable in this session and are not fabricated above; the repeat-correspondent signal is therefore unclear and should be re-run against Assignment Center.

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (or https://assignment.uspto.gov/patent/index.html), search by patent number 6141749.

Notation carried forward from the summary section: Unified Patents lists the priority date as 1997-09-11 while Google Patents and the EP family (EP 98306984.0) say 1997-09-12; I continue to treat 1997-09-12 as correct.

I'll search for the specific patent and analyze its cited prior art references.

Let me gather details on the key cited references individually.

Prior-Art Analysis — US 6,141,749 B1

Sourcing caveat (read first)

I could not open USPTO PatentCenter / PAIR / the USPTO full‑text search in this session. My attempts at USPTO‑domain endpoints did not return usable record text. What I can ground this analysis on is:

• the authoritative full text you supplied (Google Patents mirror of US6141749A), including its face‑of‑patent "Patent Citations (12)" and "Non‑Patent Citations (4)" tables; and
• corroborating web records for US5842040, US5848233, EP0743777A2 / US5878231, EP0856974A2 / US6173364, and US5842040's WO counterpart retrieved this session.

Where a detail is not confirmed by one of those, I say so. I did not auto‑correct any patent number: the table below is reproduced literally from the '749 citation list, including its duplicate/odd entries (e.g., the two Storage Technology citations US5842040 and WO9749038 covering the same subject matter, and the two Cadix/Novell‑style entries).

Note also: because the application was filed 1997‑09‑12, pre‑AIA 35 U.S.C. §§ 102(a), (b), and (e) govern. That matters for the dates below — several cited references are §102(e) art only (their US filing predates 1997‑09‑12, but they published/granted after it).

---

The claims that must be accounted for (cross-reference, not repeated)

Per the previously generated summary, claims 1, 5, 9, 13 are parallel method / apparatus / means‑plus‑function / programmed‑system forms, each requiring three things:

1. cache storing (i) a result of applying at least a portion of a rule set to a given packet and (ii) at least one hardware address associated with that packet;
2. use of that stored result for a subsequent similar packet, retrieved by a session key; and
3. determining the destination of the subsequent packet using the stored hardware address.

Dependent claims 2/6/10/14 add "bypass rule processing"; 3/7/11/15 narrow "similar" to same session; 4/8/12/16 narrow the hardware address to a LAN address.

The linchpin for §102 purposes is element (3) — the cached hardware address used for destination resolution. I flag this up front because it drives the conclusion below.

---

Master table of cited references

Dates are as listed on the '749 face (priority date / publication‑or‑grant date). "§102 posture" uses pre‑AIA subsections.

---

Reference‑by‑reference detail

1. US 5,349,642 — Kingdon / Novell, Inc.

• Citation: U.S. Patent 5,349,642, "Method and apparatus for authentication of client server communication," Novell, Inc. Priority 1992‑11‑03; granted 1994‑09‑20. §102(b) (more than one year before the 1997 filing).
• Description: Client‑server authentication scheme using challenge/response credentials.
• §102 analysis vs. claims 1‑16: No anticipation. It does not disclose packet filtering at all, let alone caching rule‑processing results keyed by a session key or storing a hardware address for destination resolution. Background/authentication‑context only. Potentially anticipates: none of claims 1–16.

2. WO 97/00471 A2 — Check Point Software Technologies Ltd.

• Citation: PCT publication WO 97/00471 A2, "A system for securing the flow of and selectively modifying packets in a computer network." Priority 1993‑12‑15; published 1997‑01‑03. §102(a) (published before the 1997‑09‑12 filing, assuming no earlier invention date).
• Description: The Check Point FireWall‑1 lineage — inspection of packets against a security policy, with selective modification. This is the reference most associated with "stateful inspection."
• §102 analysis vs. claims 1‑16: This is the most thematically relevant reference for the general stateful‑inspection idea, so it is the one to watch for claims 1/5/9/13. However, in the form cited I cannot confirm it discloses (i) a cache storing the rule‑set result together with a hardware address and (ii) using that stored hardware address to determine the subsequent packet's destination. Absent that, it does not anticipate; it is at most §103 material against the "statefulness" concept. Potentially anticipates: claims 1/5/9/13 in theory; in practice I cannot verify the hardware‑address element.

3. US 5,699,513 — Feigen et al. / Motorola, Inc.

• Citation: U.S. Patent 5,699,513, "Method for secure network access via message intercept." Priority 1995‑03‑31; granted 1997‑12‑16. §102(e) (US filing 1995‑03‑31, before applicant's filing).
• Description: Intercepting a message/connection and applying an access‑control decision at an intermediary.
• §102 analysis vs. claims 1‑16: Discloses interception and access control, but not caching of rule results or cached hardware addresses. Relevant only as context for the "session" concept in dependent claims 3/7/11/15. Potentially anticipates: none of claims 1–16.

4. EP 0 743 777 A2 / US 5,878,231 — Baehr et al. / Sun Microsystems, Inc.

• Citation: EP 0 743 777 A2, "System for packet filtering of data packets at a computer network interface." Priority 1995‑05‑18; published 1996‑11‑20. US sibling US 5,878,231 (filed 1997‑02‑04; granted 1999‑03‑02) is §102(e) art.
• Description (confirmed this session): A dedicated screening computer with three port types (private / public / proxy networks). Packets are "filtered based upon their contents, state information and other criteria, including their source and destination," and actions can include allowing/dropping or altering the IP address and forwarding to a proxy host. Packet dispositions can be taken "with or without alteration of their data, IP address, etc."
• §102 analysis vs. claims 1‑16: The explicit reliance on state information and on altering/forwarding based on destination makes this the most constructive of the general firewall references. Still, it screens each packet against content/state; it does not disclose a cache of rule‑processing results indexed by a session key and holding a hardware address used for downstream destination determination. Potentially anticipates: claims 1/5/9/13 only if "state information" were stretched to equal the claimed cache — I do not believe it supports that. Best characterized as §103.

5. WO 97/02734 A2 — Cabletron Systems, Inc.

• Citation: WO 97/02734 A2, "Internet protocol (IP) work group routing." Priority 1995‑07‑12; published 1997‑01‑30. §102(a).
• Description: IP workgroup routing; network‑layer forwarding/grouping.
• §102 analysis vs. claims 1‑16: Routing, not firewalling; no rule‑result cache or hardware‑address caching. Potentially anticipates: none of claims 1–16.

6. US 5,706,427 — Tabuki / Cadix Inc.

• Citation: U.S. Patent 5,706,427, "Authentication method for networks." Priority 1995‑09‑08; granted 1998‑01‑06. §102(e).
• Description: Network authentication of users/sessions.
• §102 analysis vs. claims 1‑16: Authentication only; no packet‑filter rule caching, no session key retrieval, no hardware address. Potentially anticipates: none of claims 1–16.

7. US 5,826,014 — Coley et al. / Network Engineering Software

• Citation: U.S. Patent 5,826,014, "Firewall system for protecting network elements connected to a public network." Priority 1996‑02‑06; granted 1998‑10‑20. §102(e).
• Description: A firewall protecting network elements from a public network.
• §102 analysis vs. claims 1‑16: Generic firewall context. No disclosure of rule‑result caching or cached hardware address. Potentially anticipates: none of claims 1–16.

8 & 9. WO 97/49038 A1 and US 5,842,040 A — Storage Technology Corp. (same subject matter)

• Citations: (a) WO 97/49038 A1, "Policy caching method and apparatus for use in a communication device," priority 1996‑06‑18, published 1997‑12‑24; (b) U.S. Patent 5,842,040 A, "…based on contents of one data unit in a subset of related data units," priority 1996‑06‑18, granted 1998‑11‑24. The US patent is §102(e) art.
• Description (confirmed this session): A communication device that caches policy decisions so that a previously computed policy result can be reused for a related data unit rather than re‑applying the policy. Current ownership/listing: Juniper Networks Inc.
• §102 analysis vs. claims 1‑16: This is the single most material reference for the caching aspect — it is the closest thing on the face of the patent to "store the result of applying at least a portion of a rule set … and reuse it." That maps to claim 1 element (1) and to the "bypass rule processing" dependent claims 2/6/10/14. But it does not disclose the '749's session‑key retrieval mechanism, and I find no teaching of a stored hardware address used to determine destination. So it does not anticipate in full. Potentially anticipates (element‑level only): claims 1/5/9/13 and 2/6/10/14 partial; no complete anticipation.

10. US 5,828,833 — Belville et al. / Electronic Data Systems Corp.

• Citation: U.S. Patent 5,828,833, "Method and system for allowing remote procedure calls through a network firewall." Priority 1996‑08‑15; granted 1998‑10‑27. §102(e).
• Description: Enabling RPC traffic across a firewall (typically via a controlled relay/proxy).
• §102 analysis vs. claims 1‑16: Deals with the proxy/relay aspects of firewall design (spec's aspects 3–5), not the claimed rule‑result cache. Potentially anticipates: none of claims 1–16.

11. US 5,848,233 — Radia et al. / Sun Microsystems, Inc.

• Citation: U.S. Patent 5,848,233, "Method and apparatus for dynamic packet filter assignment." Priority 1996‑12‑09; granted 1998‑12‑08. §102(e).
• Description (confirmed this session): An Access Network Control Server (ANCS) plus a Services Management System (SMS) that detect an event associated with a client system, select filtering rules based on the event type, and establish/replace a packet filter (e.g., per‑user filtering profiles) for packets originating at that client — used in ISP/cable‑modem contexts with dynamically assigned IP addresses.
• §102 analysis vs. claims 1‑16: This is the closest reference to the dynamic‑rules aspect (spec aspect 4), but that aspect is not recited in the '749 claims — those claims are all directed to the stateful‑cache aspect. No rule‑result cache, no session key, no hardware address. Potentially anticipates: none of claims 1–16. (Note: this reference also appears repeatedly in later reexam/IPR papers as the primary "individualized rule set" prior art for other Lucent/AT&T family patents — see the PTAB documents returned this session — but that is a different patent family, not the '749.)

12. EP 0 856 974 A2 / US 6,173,364 — Zenchelsky et al. / AT&T Corp.

• Citations: EP 0 856 974 A2, "Session cache and rule caching method for a dynamic filter," priority US 08/783,180 filed 1997‑01‑15, published 1998‑08‑05. US counterpart US 6,173,364 B1 (granted 2001‑01‑09). As a US filing predating 1997‑09‑12, the US counterpart is §102(e) art. The EP publication itself (1998‑08‑05) is NOT §102(a)/(b) art.
• Description (confirmed this session): A cache for a network filter whose rule bases change dynamically. Each cache entry stores an action derived from a rule base, keyed by the packet's identification data (explicitly a 5‑tuple: src addr, dst addr, src port, dst port, protocol), plus a rule‑base indicator and version number. On a packet, the filter searches the cache first; a subsequent packet in the same "message" with the same 5‑tuple gets the cached action without searching the rule bases ("subsequent packets essentially benefit from the search done for the first packet … This saves processor time and increases the throughput of the filter"). The reference even frames the problem as redundant re‑searching because all packets in a message share the same 5‑tuple.
• §102 analysis vs. claims 1‑16: This is the strongest reference against the independent claims. It discloses, at the element level: caching a rule‑processing result, keying/retrieving by a session key (5‑tuple), bypassing rule‑base search for subsequent same‑session packets, and the "same session" narrowing. Those map to independent claims 1/5/9/13 (elements 1–2), and directly to dependent claims 2/6/10/14 and 3/7/11/15. What it does not disclose is the '749's distinguishing third element: storing a hardware (e.g., LAN) address in the cache and using it to determine the subsequent packet's destination — the subject of claims 1(iii)/4/8/12/16. Because that element is absent, EP'974/US'364 does not fully anticipate any claim. Potentially anticipates: element‑level for claims 1/5/9/13, 2/6/10/14, 3/7/11/15; no complete §102 anticipation of any claim.

---

Non‑patent citations

Check Point FireWall‑1, OPSEC Open Specification, v1.01 (Nov. 1998)

• Published Nov. 1998 — after the 1997‑09‑12 filing date. Unless a pre‑filing edition can be established, this is not prior art to the '749 patent under any §102 subsection. No §102 anticipation.

EliaShim Ltd. press release, "Announces CVP‑Compliant Anti‑Virus Plug‑In for Check Point FireWall‑1" (Feb. 17, 1997)

• A printed publication dated Feb. 17, 1997, i.e., before the filing date → §102(a) posture. Its subject matter — an anti‑virus plug‑in to a firewall — does not touch rule‑result caching, session keys, or hardware addresses. Potentially anticipates: none of claims 1–16.

---

Bottom line

Most relevant prior art for US 6,141,749:

1. US 5,842,040 A / WO 97/49038 A1 (Storage Technology) — policy caching. Closest on the concept of caching/reusing a computed policy (rule) result, i.e., the "bypass rule processing" idea in dependent claims 2/6/10/14.
2. EP 0 856 974 A2 / US 6,173,364 (AT&T) — session cache & rule caching for a dynamic filter. Closest on the mechanics of the independent claims: cache keyed by a 5‑tuple session key, cache‑first lookup, rule‑base bypass for subsequent same‑session packets.
3. EP 0 743 777 A2 / US 5,878,231 (Sun) — packet screening with state information.
4. WO 97/00471 A2 (Check Point) — stateful inspection / selective packet modification.
5. US 5,848,233 A (Sun, Radia) — dynamic packet‑filter assignment (relevant to the spec's dynamic‑rules aspect, but not recited in the '749 claims).

Anticipation conclusion (candid): On the record I can access, none of the twelve cited patent references fully anticipates any of claims 1–16. Every independent claim requires a cached hardware address used to determine the subsequent packet's destination (further narrowed to a LAN address in claims 4/8/12/16), and I find no teaching in any cited reference of storing a hardware address in the rule‑result cache for that purpose. The two Storage Technology and AT&T references are closer to the caching and session‑key elements and would be the natural §103 combinations, but element 3 appears to be the point of novelty that distinguishes the '749 claims.

Contradiction to flag: The previously generated summary's "Family" line lists EP0909074A1 (multiple domain support) as a family member. That family member is not among the references cited on the '749 face (the face lists EP 0 856 974, a different EP application — the AT&T one). Treat EP0909074 and EP0856974 as distinct; the prior‑art tables above use the literal '749 citation list. Also note the two Storage Technology citations (US5,842,040 and WO97/49038) are the US and PCT counterparts of a single disclosure, not two independent references.

What I could not verify: I have no direct USPTO PatentCenter/PAIR confirmation of the citation list beyond the Google‑Patents mirror of the '749 front page; and I have not inspected the full text of US5,348,642, US5,699,513, US5,706,427, US5,826,014, US5,828,833, WO97/00471, WO97/02734, or US5,842,040 line‑by‑line in this session. My "no anticipation" findings for those rest on the titles/abstracts and the confirmed content of the AT&T, Sun, and Storage Technology references — so treat the per‑reference §102 mappings as element‑level hypotheses suitable for a full‑text novelty search, not as a certified invalidity opinion.

I'll ground this analysis in the actual prior-art references listed on the patent page before constructing the §103 combinations.

Now let me examine the remaining cited references, especially for the hardware-address limitation that distinguishes every independent claim.

Obviousness Analysis — US 6,141,749 B1 under 35 U.S.C. § 103

Framing notes before the analysis:

1. Date discrepancy (flagging, not auto-correcting): the task header states "Current Date: April 26, 2026," whereas my operating rules state today is 2026-09-30. I treat 2026-09-30 as operative for "today's" statements; the difference does not affect any legal conclusion below.
2. Governing law: Application 08/928,794 was filed 1997-09-12, so pre-AIA § 102/§ 103 govern. KSR Int'l v. Teleflex still supplies the obviousness standard; the pre-AIA "teaching, suggestion, or motivation" requirement is satisfied by any articulated rational underpinning (In re Kahn; MPEP 2143/2144).
3. Scope: I use the 12 references on the face of the patent ("Patent Citations") plus the 4 NPL items, as instructed. The "Cited By" lists (86 / 189 entries, e.g., Nicira US 9,825,913, VMware US 11,899,594) are later-published documents and therefore not § 102 prior art against a 1997 filing — I exclude them except where a family member predates the filing.

---

1. What must be met — the only limitations that matter

All 16 claims are drawn to the "stateful packet filtering" aspect (summary §"Plain-language overview"). Claims 1/5/9/13 are parallel (method / apparatus / means-plus-function / programmed processor) and require exactly three things:

Claims 2/6/10/14 add "bypass rule processing"; claims 3/7/11/15 define "similar" as "same network session"; claims 4/8/12/16 narrow the hardware address to a LAN address. The specification confirms the intended reading: FIG. 4's cache contains "hardware address information," and step 505 determines the destination interface "using the local area network (LAN) address of the packet." So (B)+(D) = cache the packet's link-layer address alongside the rule result, and use it for forwarding/destination determination on subsequent packets.

Every ground below must therefore supply (B) and (D); (A) and (C) are essentially admitted art.

---

2. Reference-by-reference capability map (verified vs. unverified)

---

3. Proposed grounds of rejection

Ground 1 (primary) — US 5,842,040 in view of WO 97/02734 (US 5,751,971)

• Claims 1, 5, 9, 13: US 5,842,040 teaches storing, for a session's first PDU, the cached policy identification information (= the rule-processing result) and applying it to subsequent related PDUs (elements A, C); it expressly names a firewall as the host device. WO 97/02734 teaches maintaining a cached host table associating host addresses with network interfaces and physical addresses, and using the physical-address/interface association to control where a packet is forwarded (elements B, D). The two references are analogous art: both are network forwarding/access-control devices operating on packet headers at the same point in the stack.
• Claims 2, 6, 10, 14 (bypass): met by US 5,842,040 on its face — cached policy is applied "without requiring a complete analysis of the other related PDUs."
• Claims 3, 7, 11, 15 (same session): met by US 5,842,040's "group of related PDUs… (called a session)."
• Claims 4, 8, 12, 16 (LAN address): met by WO 97/02734's physical-address (i.e., link-layer) association.

Ground 2 — EP 0 856 974 A2 (via US 6,173,364, AT&T) in view of WO 97/02734

• AT&T supplies a true session cache of rules: search the cache by a key derived from the received packet's identification data; on a miss, search the rule base, carry out the rule, copy it into the cache for subsequent packets (A, C; claim 2's bypass; claim 3's session identity, since the key is the 5-tuple). Adding the Cabletron physical-address/interface association to the cache entry supplies (B) and (D). This ground is attractive because "session cache" is literally the claim's mechanism.
• § 102 caution: the EP A2 publication (1998-08-05) postdates the '749 filing and is not prior art as a printed publication. The ground must be pleaded on the US counterpart application 08/783,180, filed 1997-01-15, which issued as US 6,173,364 — qualifying § 102(e) art. Confirm the US disclosure supports the same text before relying on it.

Ground 3 — EP 0 743 777 A2 (Sun screen) in view of US 5,842,040

• Sun supplies the firewall that filters on "contents, state information and other criteria," passes/drops/alters addresses, and consults prior state for subsequent packets (C). US 5,842,040 supplies explicitly caching that determination for the session's related packets (A, C, bypass). Ground 3 is the better § 102 posture (Sun's EP published 1996-11-20 is solidly prior art), but it is weaker on (B)/(D) and must borrow the hardware-address teaching from WO 97/02734, making it a three-reference combination.

Ground 4 — any primary above in further view of US 5,848,233 (Radia)

Radia supplies the reason a firewall would maintain per-user/per-session cached state at all: individualized filtering profiles that are re-provisioned as users log in and out. It is best used as motivation/context, not as a teaching of the cache (Radia's mechanism is reconfiguration, not caching).

---

4. Motivation to combine (the KSR/MPEP rationales)

A § 103 challenge stands or falls on articulated reasons. The strongest, in order:

1. Known technique applied to a known device ready for improvement, predictable result (MPEP 2144.04(V)/2143(A)). US 5,842,040 identifies the exact problem the '749 states it solves — per-packet rule evaluation throttles throughput — and states the exact solution the '749 uses (cache the result of analyzing one packet of a session, apply it to the rest). Applying that to a filter with per-user rule sets (Radia) is a predictable efficiency gain.
2. Simple substitution / use of a known element for its known function. Both StorageTek's and AT&T's caches are keyed and indexed by packet identification data. Caching a link-layer address in the same entry and using it to resolve the egress interface is the standard ARP/forwarding-table function taught by WO 97/02734; substituting/adding it is the use of a known element for its known purpose.
3. Design incentive / recognized need. A firewall that resolves the "pass" action must still emit the frame; re-resolving the hardware address on every packet of a session defeats the very caching the primary reference is designed to achieve. The '749 claim's own file-wrapper describes exactly this (step 505: destination interface determined from "the LAN address of the packet"). The efficiency motivation is thus explicit and non-hindsight.
4. Same field, same problem, complementary teachings. US 5,842,040 names firewalls; WO 97/02734 addresses access control on routed IP traffic; EP 0 743 777 is a firewall; all three are packet-forwarding access-control arts — "reasonably pertinent to the particular problem" (KSR; In re Bigio).
5. Predictability / no teaching away. Nothing in the references disparages storing link-layer addresses with connection state; Ethernet/Token-Ring media are addressed by hardware address by definition, so caching an L2 address is an inevitable bookkeeping step.

---

5. Dependent-claim conclusion

Contingent on Ground 1 or 2, dependent claims 2–4 (and mirrors 6–8, 10–12, 14–16) add no independent inventive weight: claim 2's bypass is the express purpose of US 5,842,040; claim 3's "same session" is the express grouping criterion of US 5,842,040 and the very name of AT&T's "session cache"; claim 4's "LAN address" is the physical-address teaching of WO 97/02734 and is otherwise a plain consequence of L2 framing. Under KSR, these are "predictable variation[s]" of the combined teachings.

For the means-plus-function claim 9, § 112(f) construction maps "means for storing/utilizing/determining" to the disclosed domain support engine, cache memory, NIC, and processor. Prior-art network processors of the US 5,842,040 / WO 97/02734 / Radia type disclose that identical structure, so the § 103 conclusion is unchanged (the §112 indefiniteness concern already flagged in the earlier summary affects construction, not separate patentability).

---

6. Weaknesses, counter-arguments, and integrity flags

1. The examiner allowed over these very references. US 5,842,040, WO 97/02734, EP 0 856 974, EP 0 743 777, US 5,848,233 and US 5,828,833 are all on the face of the '749 patent. That strongly implies the examiner did not read any of them as teaching caching a hardware address and using it to determine the subsequent packet's destination. Any petitioner must therefore attack that single link head-on (e.g., with the Cabletron physical-address/interface material or an admission in the '749 specification that LAN-address destination determination was conventional — note step 505 is described matter-of-factly, not as novel).
2. None of the cited references' caches store a hardware address. StorageTek caches "policy identification information"; AT&T caches a 5-tuple-keyed rule + version; Sun caches state information and alters IP addresses. The hardware-address link is supplied only by WO 97/02734, and see flag 3.
3. ⚠️ Verification gap on WO 97/02734. The Google Patents abstract I retrieved confirms the work-group/interface-assignment concept but not the host-table-cache or physical-address limitations. Those I observed in US 6,249,820, which was filed 1998-05-06 — after the '749 filing. Before pleading Ground 1/2 on this element, the US 5,751,971 specification (filed 1995-07-12) or the WO 97/02734 pamphlet text must be confirmed to disclose caching a host's physical address and using it to route. I could not verify that in this session; treat Ground 1's (B)/(D) mapping as unconfirmed.
4. § 102 date defects to plead around (do not auto-correct):
   • EP 0 856 974 A2 published 1998-08-05, ~11 months after the '749 filing → not § 102(a)/(b) art; use US 6,173,364 (app. 08/783,180, filed 1997-01-15).
   • WO 97/49038 published 1997-12-24 → not art; use US 5,842,040 (filed 1996-06-18).
   • WO 97/00471 (1997-01-03) and WO 97/02734 (1997-01-30) are § 102(a)-class art only — they are inside the one-year grace period, and are the references most exposed to a Rule 131/antedating attack. The AT&T litigation record shows exactly this tactic succeeding in the same technological family: in IPR2014-00786 (AT&T's US 6,826,694) the patent owner swore behind 1997-dated references (Rubin, Cunningham) with a June 1996 reduction to practice and third-party corroboration (hearing transcript). Coss/Majette/Sharp (Lucent) filed 1997-09-12 and could plausibly attempt the same against WO 97/02734 and WO 97/00471.
5. References I could not verify in this session (listed in the Patent Citations but with no text retrieved): US 5,826,014 (Coley), US 5,349,642 (Kingdon), US 5,706,427 (Tabuki), US 5,828,833 (Belville), WO 97/00471 (Check Point). I make no contention about what they disclose. If the Check Point WO 97/00471 specification is as broad as its title ("securing the flow of and selectively modifying packets"), it may independently disclose stateful connection tracking with address modification, which would strengthen Grounds 1–3; that must be verified against the document.
6. Practical posture (carried over from the prior summary): the '749 term expired 2017-09-12; its claims are unassertable for prospective infringement. A 2026 § 103 analysis is therefore most likely needed for a retrospective validity opinion, an ex parte reexamination, or an expired-patent/IPR-style challenge tied to past Sound View Innovations assertions — not for a live CAFC appeal. I found no docket tying 6,141,749 itself to a 2026 appeal.

---

7. Bottom line

• Elements (A) and (C) — caching a rule-set result for a session's packet and reusing it, keyed by a session key, to bypass rule processing for subsequent packets — are squarely and redundantly disclosed by US 5,842,040 (Storage Technology, filed 1996-06-18) and US 6,173,364 / EP 0 856 974 (AT&T, filed 1997-01-15), both § 102(e) art. Claims 2 and 3 (and mirrors) are essentially anticipated in substance by those disclosures.
• Elements (B) and (D) — the cached hardware address and its use to determine the destination — are the only genuine point of novelty, and the only cited reference capable of supplying them is WO 97/02734 (Cabletron) / US 5,751,971, whose disclosure on this point I could not confirm.
• Most probable § 103 grounds: (1) US 5,842,040 + WO 97/02734; (2) US 6,173,364 + WO 97/02734; (3) EP 0 743 777 + US 5,842,040 (+ WO 97/02734). Each has a credible articulated motivation (throughput by session-level caching; L2 address resolution is the known way a forwarding device picks an egress port).
• Confidence: moderate-to-high that claims 1–16 are obvious if the Cabletron reference is shown to cache a physical/hardware address used for forwarding; low that the patents are obvious on the face of the printed citations alone, because the examiner allowed over all of them and none of the others' caches stores a hardware address.